Received: 26 July 2021 Accepted: 6 January 2022 Published online: 18 February 2022

DOI: 10.1002/csc2.20710

Crop Science
ORIGINAL RESEARCH ARTICLE
Crop Breeding & Genetics

Control of aflatoxin using atoxigenic strains and irrigation

management is complicated by maize hybrid diversity

Jacob J. Pekar1 Seth C. Murray1 Thomas Isakeit1 Luke S. Pruter2

Nancy J. Wahl1 Michael J. Brewer2

1 Texas A&M Univ., Texas A&M AgriLife

Research, College Station, TX 77843-2474, Abstract

USA
2Entomology Program, Texas A&M
AgriLife Research and Extension Center,
10345 State Hwy. 44, Corpus Christi, TX
78406, USA
Correspondence
Jacob Pekar, Texas A&M Univ., Texas A&M
AgriLife Research, College Station, TX
77843-2474, USA.
Email: jacob.pekar@outlook.com
Assigned to Associate Editor Candice Hirsch.
Funding information
National Institute of Food and Agriculture,
Grant/Award Number: 2014-6804-21836;
Texas corn producers board; Agricultural
Research Service; Aflatoxin Mitigation Cen-
ter of Excellence; Texas AgriLife Research
Mycotoxins produced by the fungus Aspergillus flavus are harmful to humans and
animals that feed on contaminated products. This study was conducted to determine
if there are any interactions between applications of an atoxigenic A. flavus, irriga-
tion regimes, locations, and hybrids that affect contamination of maize (Zea mays L.).
Total genotypic variation was 3.8% for transformed aflatoxin data, suggesting that the
diverse hybrids used were a minor to moderate driver of aflatoxin variation in this
experiment. Location (Loc) × year × irrigated (Irr) × genotype (Gen) for transformed
aflatoxin explained 3.3% variation, which is useful in selecting reduced aflatoxin-
susceptible genotypes in differing environments. Variation from atoxigenic inocula-
tion (Inoc) explained 2.7% variation and reduced aflatoxin accumulation by 11% on
average. Location × Inoc explained 2.1% of variation, supporting the effectiveness of
atoxigenic applications dependent on different locations. All effects containing Inoc
× Gen totaled .45%, suggesting that genotypes could have a small influence on effec-
tiveness of atoxigenic applications. These are critical components to understand that
can aid a breeder in selecting aflatoxin-tolerant germplasm.

1 INTRODUCTION 2005), especially in developing countries that lack infrastruc-

Aflatoxin is a harmful carcinogenic mycotoxin produced by
Aspergillus flavus Link:FR, which limits the marketability of
maize (Zea mays L.) grain and reduces the economic value
for producers. Actual economic losses are difficult to measure
but are thought to be around US$163 million per year in the
United States in maize only and up to $500 million annually
in peanuts (Arachis hypogaea L.) and other crops (Wu, 2015).
Documented or suspected cases of acute aflatoxin poisoning
are numerous throughout the world and result in liver dam-
age, intestinal bleeding, cancer, and even death (Lewis, et al.,
Abbreviations: BLUP, best linear unbiased predictor; DTA, days to
anthesis; DTS, days to silking; EHT, ear height; PHT, plant height.
ture to test for contamination and allow contaminated maize
to flow freely in local trade. The effects of chronic exposures
to aflatoxins are even more challenging to test. More than 100
countries have some regulations of aflatoxin concentrations in
feed or foods (Wu & Guclu, 2012). In the United States, afla-
toxin is regulated by the U.S. Food and Drug Administration
for human consumption with an upper limit of 20 (ng g−1 ),
while the maize upper limit of 300 (ng g−1 ) can be used for
finishing beef, with intermediate, lower limits for swine and
poultry feed (Stoloff, et al., 1991).
It remains unclear why A. flavus makes aflatoxin, but it may
prevent insect predation of the fungus (Drott, et al., 2017;
Gqaleni, et al., 1997). Preharvest colonization of maize and

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original
work is properly cited.
© 2022 The Authors. Crop Science published by Wiley Periodicals LLC on behalf of Crop Science Society of America.

Crop Science. 2022;62:867–879. wileyonlinelibrary.com/journal/csc2 867

868 Crop Science
the subsequent production of aflatoxins are associated with,
and likely a result of, an increase in physiological stresses of
crop production (Klich, 2007). High daytime and nighttime
temperatures, along with occurrences of drought and insect
pressure, have been shown to increase the occurrences of afla-
toxin contamination (Abbas, et al., 2002). To reduce pathogen
pressure and toxin accumulation, producers decrease stresses
through cultural practices and management when econom-
ical. It is well known that changing management effects
within environments, such as optimal fertilization and irri-
gation effects, can affect aflatoxin accumulation. Generally,
procedures that reduce plant stress reduce aflatoxin (Payne &
Widstrom, 1992; Robens & Cardwell, 2003). In recent years,
atoxigenic strains of A. flavus have become an additional man-
agement tool to reduce aflatoxins (Abbas, et al., 2011). These
atoxigenic strains have a defective mutation in the aflatoxin
pathway. The spores of atoxigenic strains are then spread
throughout producers’ fields on a nutritive carrier of grain—
wheat (Triticum aestivum L.) or sorghum [Sorghum bicolor
(L.) Moench]—and by sheer numbers, outcompete the native
aflatoxin-producing strains for colonization sites in the maize
kernels. However, while this competitive exclusion decreases
aflatoxin by 40–95% (Abbas, et al., 2011; Brown, et al., 1991;
Isakeit, et al., 2010; Isakeit, et al., 2011), this inoculation typ-
ically increases ear rot of the ear tip by the A. flavus fungus.
Plant breeding is an important component of the integrated
pest management approach to decrease aflatoxin and A. flavus
(Brown, et al., 2011; Williams, 2006). Breeding for decreased
susceptibility by selecting for heritable segregating traits such
as tighter, thicker, and closed husk cover or insect resistance
(native or transgenic) can reduce toxin accumulation (Wid-
strom, et al., 2003). Hybrid and inbred selections in maize
have been shown to reduce preharvest aflatoxin accumulation
by up to 90% (Abbas, et al., 2011; Brown, et al., 1991; Mur-
ray, et al., 2019; Pekar, et al., 2019; Wahl, et al., 2016). A
major challenge of breeding inbred lines for aflatoxin or yield
is that inbred lines appear to be less robust across environ-
ments (i.e., experience more genotype × environment inter-
actions) and experience more stress, leading to higher levels
of contamination, in a way that is masked by hybrid vigor
in hybrids (Cole, et al., 2009; Li, et al., 2018; Schnell &
Becker, 1986). To uniformly evaluate the susceptibility of
maize genotypes to A. flavus and aflatoxin accumulation, arti-
ficial inoculation must be used to induce infection. There are
two major classes of A. flavus inoculation methods: wound-
ing and nonwounding (Tucker, Jr., et al., 1986). Wounding
inoculation techniques offer consistent infections that mimic
insect or mechanical damage, bypassing resistance expressed
in physical and physiological traits such as husk cover or
husk tightness, natural insect resistance, pericarp thickness,
and erect or nonerect (drooping) ears. Nonwounding meth-
ods, mimicking the approach used by producers to apply atox-
igenic strains, do not bypass physical or physiological bar-
PEKAR ET AL.
Core Ideas
∙ Environmental and agronomic interactions for
breeding aflatoxin tolerant maize hybrids were
investigated.
∙ Interactions between genotype, environmental,
agronomic, and inoculation methods and how these
interactions could lower aflatoxin accumulation
were analyzed.
∙ The possibility of synergistic and antagonistic
interactions for aflatoxin tolerance in maize was
investigated.
riers to infection tolerances, and they evaluate susceptibility
in a way that is more relevant to producers (Williams, et al.,
2002; Windham, et al., 2009). The different types of inoc-
ulation methods have various advantages and disadvantages
(Williams, et al., 2013). In this research, we chose to use a
nonwounding technique, ground kernel inoculation, because it
more closely mimics natural inoculation conditions and atox-
igenic A. flavus deployment. This technique also reduces the
labor needed, allowing the evaluation of a larger number of
genotypes and replicates for aflatoxin accumulation.
If the aflatoxin-reducing effects of improved hybrids
(including insect resistance traits), improved agronomic man-
agement, and deployment of A. flavus atoxigenic strains
worked synergistically, aflatoxin should be nearly eliminated
from producers’ fields when these approaches are combined;
however, no investigation into these interactions has yet been
attempted. It is possible that these effects could be indepen-
dent or antagonistic as opposed to synergistic and cumulative,
which would be important in developing systems-level recom-
mendations. A comprehensive investigation into interactions
between genotype × environment × management × atoxigenic
applications is labor intensive, requiring a large number of
replications to ensure enough statistical power to test the many
interactions within a model.
Perhaps due to the complexity of looking at such interac-
tions, most control methods have only been investigated in
isolation. However, it has been previously shown that certain
biological agents, such as rhizobia and arbuscular mycorrhizal
fungi, have host-genetic specificity that enables specific plant
genotypes to allow host fungal interaction (Kloepper, 1996;
Smith & Goodman, 1999). In the case of maize, genetics
plays an important role in host susceptibility to A. flavus
(Abbas, et al., 2011; Brown, et al., 1991; Murray, et al., 2019;
Pekar, et al., 2019; Wahl, et al., 2019), but this has typically
been measured using aflatoxin accumulation. A. flavus col-
onization is necessary for the production of aflatoxin but is
not directly correlated to aflatoxin production (Henry, et al.,
PEKAR ET AL.
2009). A. flavus load itself is difficult to impossible to measure
in plants even with tools like reverse transcription-polymerase
chain reaction (Mideros et al., 2009). Without genotype ×
atoxigenic applications data available, it has been necessary
to assume that atoxigenic strains of A. flavus would act in a
similar manner to the toxigenic strains. However, it is reason-
able to hypothesize that certain plant phenotypes or unseen
genetic interactions could induce preferential susceptibilities
to infection of either the atoxigenic or the native toxigenic
strains. This interaction could potentially be exploited for
gains in the synergistic reduction of aflatoxin accumulation
through genotype × atoxigenic applications. Exploiting this
interaction could allow producers to apply atoxigenic strains
in hybrids that primarily host toxigenic strains. Similarly,
tolerance to water and nutrient stress affects aflatoxin produc-
tion (genotype × management; Jones, et al., 1981). While we
hypothesize that correctly managing genotype × atoxigenic
applications × management could decrease aflatoxin by
more than 90%, we also hypothesize that ignoring these
interactions could result in aflatoxin levels being higher than
expected.
The specific objective of this study were to a) evaluate
the aflatoxin accumulation in diverse commercial and exper-
imental hybrids known to have varying levels of suscepti-
bility to aflatoxin; b) evaluate the sources of variance from
the main effects of environment, management (irrigation vs.
dryland), and atoxigenic applications (atoxigenic and toxi-
genic inoculated vs. toxigenic only); and finally c) evalu-
ate the proportion of variance from the interactions between
main effects to determine synergistic or antagonistic relation-
ships between aflatoxin control measures. This study will sug-
gest how closely researchers, agricultural companies, and pro-
ducers should manage these control measures for real-world
reductions in aflatoxin accumulations.
2 MATERIALS AND METHODS
Beginning in 2014, irrigated (Irr) and dryland (Dry) trials
were initiated in Corpus Christi, TX, and College Station, TX.
Experimental design consisted of a modified split-plot ran-
domized complete block design of 32 entries in 2014, includ-
ing local commercial checks, and 28 entries the following
years (2015, 2016, and 2017) in two locations (Corpus Christi,
College Station), including local commercial checks, each
having two management approaches of irrigated and dryland
(Irr, Dry). The number of checks varied due to seed availabil-
ity from 11 checks in 2014, 16 checks in 2015 and 2016, and
8 checks in 2017. Planting populations and fertilizer applica-
tions in College Station were 81,000 plants per hectare in the
irrigated trial with a fertilization rate of 134 kg/ha on nitrogen
and 61 kg/ha of phosphorous while the dryland had a fertil-
ization rate of 90 kg/ha nitrogen and 61 kg/ha of phospho-
Crop Science 869
rous and 64,000 plants per hectare. Corpus Christi trials were
planted at 74,000 plants per hectare in both irrigated and dry-
land. College Station and Corpus Christi had two inoculation
types (atoxigenics applied (ATOX) and no atoxigenics applied
[noATOX]), each with four replications, ensuring appropri-
ate statistical power for investigating these interactions. In the
split-split-split-plot treatments, irrigated and dryland was the
main split followed by the atoxigenic application split fol-
lowed by genotypes.
Fields were selected for uniformity to minimize residual
error from unexplained field variation. The size of the trial
was 0.70 hectares, and design was an important factor for try-
ing to reduce residual (i.e., unexplained) error. Replications
within trials were randomly allotted so that field variation was
minimized. Field variation has typically been associated with
range and row effects (also called row and column, respec-
tively) that are generally attributed to fertilizer and flood irri-
gation. In College Station, trials and replications within trials
were placed so that the furrow irrigation effects were as uni-
form as possible. This was less of a concern in Corpus Christi
due to the installation of sub-surface drip.
2.1 A. flavus inoculation
Half of the plots (ATOX) were treated with the commercial
formulation of an atoxigenic strain, Afla-Guard (Syngenta
Crop Protection, 2017), at a rate of 22.3 kg/ha just prior to
tasseling. All plots were then treated with A. flavus isolate
NRRL 3357 grown on autoclaved corn kernels. In brief, toxi-
genic inoculum was applied shortly after the majority of plots
had started silking at an amount of 6 oz. per 9.14 linear row
meters, usually around 12 d after the start of silking. Inocu-
lum was prepared from stock A. flavus isolate NRRL 3357
as previously described (Isakeit et al., 2010). Similar to Far-
fan et al. (2013), following maturity, sub-samples were taken
from the research combine. These samples were then ground
in a Romer mill (Romer Labs), and aflatoxin levels were mea-
sured via the Vicam AflaTest (VICAM).
2.2 Traits measured
Secondary phenotypic traits were measured at the College
Station location in all years for days to anthesis (DTA), days
to silking (DTS), plant height (PHT) from the ground to the
tip of the tassel, and ear height (EHT) from the ground to the
primary ear’s node attachment point. Grain yield was mea-
sured on a plot basis with a JD 3300 modified research com-
bine retrofitted with a Harvest Master classic grain gage for
total plot weight, moisture, and test weight (Juniper Systems,
1993–2017), and subsamples were taken from each plot dur-
ing harvest. Many of the same traits were taken most years in
Corpus Christi, but the ears were hand-harvested.
870 Crop Science
2.3 Statistical analysis
Statistical analysis was conducted using R (R Development
Core Team, 2010). Assuming a nonnormal distribution, as
is almost always the case for aflatoxin measurements, all
aflatoxin data were first transformed by logarithmic transfor-
mation to normalize the data (Betran, et al., 2002; Williams &
Windham, 2015). Models were then analyzed with both raw
and transformed data using the lme4 R package (Bates, et al.,
2015) and validated with JMP (JMP14, 1989–2019). Aflatox-
ins in ppb (Afppb) and transformed aflatoxin data (AfTrans)
were both reported throughout the analysis in order to offer
increased perspective on actual aflatoxin concentration.
For these analyses, it was assumed that test hybrids were
unrelated and not randomly selected; thus, repeatability was
estimated instead of heritability (Equation 1). Data were first
analyzed jointly across all years, locations, and management
practices to evaluate overall trends and then were analyzed
individually by environments (location by year). In the all-
years-combined analysis, hybrids’ genetic effects were treated
as fixed while all other factors were random. Factors included
in the model were location, year, irrigation management, and
inoculation as well as all possible interactions between these.
In order to detect the contribution of variation from all factors,
an all-random model was used for both full model and sepa-
rate year analysis. Here range and row factor effects reflect the
terminology used in furrow irrigation where water and trac-
tors drive down the row. The full model was fit as a five-level
full factorial design. All means were compared using Fisher’s
protected least significant difference (P = .05). Repeatability
(h2 ) was calculated as:
( )
GE ε
ℎ2 = 𝐺∕ 𝐺 + + (1)
𝑟 𝑟𝑒
where G, GE, and Ɛ were the variance components of geno-
type, genotype by environment interaction, and residual error,
respectively, with r as number of replications and e as number
of environments.
Correlations were compared using Pearson correlation
coefficients. Phenotypic correlations were formed on raw
data, while genotypic correlations were formed from the best
linear unbiased predictor (BLUP) genotypic estimates of each
trait extracted from the reduced model.
3 RESULTS AND DISCUSSION
To best discriminate genotypes and other treatment effects,
higher environmental aflatoxin levels are useful. Here, afla-
toxin accumulation was sufficient to discriminate genotypic
sources of variation in all years. In College Station, aflatoxin
accumulation was highest during the 2016 growing season,
reaching levels up to 940 ng g−1 for an individual sample,
PEKAR ET AL.
well above the 300 ng g−1 upper regulatory limit (Table 1).
Best linear unbiased predictions means in this same environ-
ment were 218 ng g−1 . When comparing the check hybrids
to the experimental hybrids during the College Station 2016,
which is the season with the highest accumulation, it was
found that the check group had significantly higher accrual of
aflatoxin than the experimental group (227 and 134 ng g−1 ,
respectively). In Corpus Christi, levels during the 2016 sea-
son reached levels of 1,800 ng g−1 for an individual sample
with a mean of 73 ng g−1 . In the overall analysis, combin-
ing both locations for AfTrans, the highest variation explained
was from the effect of location × year (38.3%). In AFppb, this
variation was only 6.6%. Nonnormal data, which is almost
always the case with Afppb aflatoxin data, caused an anomaly
fixed by log-transformation. The treatment for Inoc was suc-
cessful on average, as the mean of samples across all environ-
ments with the atoxigenic treatment applied was 39% lower
than untreated control (p < .001). This was substantially less
than the 75 to 90% that has been reported and suggests con-
trol factors might not have synergistic effects (Abbas, et al.,
2011; Brown, et al., 1991; Isakeit, et al., 2010; Isakeit, et al.,
2011).
Hybrid grain yield means ranged from 3.4 to 10.3 MT ha−1
in each environment. Yields were highest in College Station
2014 overall. Yield is a primary grower criterion, and the
best experimental hybrid yield was 112% of the hybrid check
average in the 2015 environment. Competitive advantage of
the experimental hybrids could be due to the nonadaptation
of commercially marketed hybrid cultivars used as checks.
The commercial checks sold in Texas have not been bred
(as inbreds or hybrids) in Texas (Murray, et al., 2019) nor have
they been specifically selected for the unique Texas environ-
ments tested (Farfan, et al., 2013). Based on 2017 data, the
Corpus Christi 2017 location flowered significantly faster at
53 DTA and 53 DTS, while the average across all recorded
years was 61 DTA and 60 DTS, respectively. Later than opti-
mal planting was conducted for all environments to maximize
stress for increased aflatoxin pressure. Later planting also
compressed flowering time among genotypes due to faster
heat unit accumulation, which was beneficial for applying
more uniform inoculations. A drawback to later planting and
compressed flowering times was that there was less variation
present to determine how flowering time might have affected
aflatoxin levels and grain yields.
Plant height and EHT in all environments had a BLUP
mean range of 179.3 to 252.2 cm and an EHT BLUP mean
range of 46.3 to 96.2 cm, suggesting variation in stress
between environments that limited or favored plant growth.
In environments with lower PHT and EHT, yield also tended
to be lower, which is commonly observed in the stressful con-
ditions of Texas (Farfan, et al., 2013; Liu & Wiatrak, 2011;
Yin, et al., 2011).
PEKAR ET AL. Crop Science 871

T A B L E 1 Summary statistics on best linear unbiased predictors (BLUPs) for transformed aflatoxin data, actual ng g-1 of aflatoxin, test weight,
yield, days to 50% anthesis, days to 50% silking, plant height, and ear height separated by location and years. Environment is a combination of
location and year for College Station (CS) and Corpus Christi (CC)

All
years CC14 CC15 CC16 CC17 CS14 CS15 CS16 CS17
Log-transformed Max 2.5 2.5 1.8 1.8 2.5 1 1.9 2.6 1.1
[Aflatoxin (ng
g−1 )+1]
Mean 1.1 2.1 0.8 1.4 1.7 0.5 0.6 1.2 0.4
Min 0.5 1.6 0.3 0.7 0.8 0.1 0.2 1.7 0.4
SD 0.3 0.2 0.4 0.2 0.4 0.2 0.4 0.3 0.2
Aflatoxin (ng Max 391.1 477.4 187.7 142.8 412.3 120.6 146.9 481.9 179.5
g−1 )
Mean 85.9 204.3 35.7 72.6 122.9 30.5 12.8 217.7 55.4
Min −30.5a 67 2.4 14.3 15.4 1.9 0.9 77.3 5.9
SD 68.1 95.4 45.1 37 96.9 30.1 27 112.3 41.2
Test weight Max 61.8 58.5 62.1 54.8 61.6 61.1 59 60.6
Mean 57.7 54.7 59.2 56.1 58.3 58.8 55.2 58.2
Min 53.3 51.1 54.4 53.4 54.8 56 51.9 55.2
SD 1.8 2 1.9 0.9 1.7 1.5 1.9 1.3
Yield (MT ha−1 ) Max 9.5 8 9.6 7.4 4.8 12.1 9.7 10.5 10.3
Mean 7.7 5.3 7.2 5.9 3.4 10.3 8.3 9 8.6
Min 3.7 3.4 5.5 5.3 2 6.6 6.8 6.8 6.7
SD 1 1.2 1 0.5 0.6 1.3 0.7 0.8 0.8
Days to anthesis Max 68.3 – – – 57.5 83.1 65.3 66.8 66
Mean 63.3 – – – 54.7 78.5 61.6 61 60.8
Min 59.2 – – – 51.9 75.4 57.1 58.1 57.2
SD 1.9 – – – 1.2 1.8 2.2 2.1 2.4
Days to silking Max 69.5 – – – 59.1 82.9 67.4 67.4 67.5
Mean 63.3 – – – 55.6 79.1 63.1 61.7 61.4
Min 59.2 – – – 53.3 75.7 57.9 58.1 57.3
SD 1.9 – – – 1.5 1.8 2.8 2.3 2.9
Plant height (cm) Max 254.9 245.5 205.2 227.3 216.3 275.1 238.6 259.7 274.2
Mean 227.8 226.8 179.3 214.5 190.8 252.2 220.2 237.3 240.3
Min 197.6 205.3 146.1 204.9 168.4 222.1 199.4 218.6 211.3
SD 11.5 11.2 15.4 5.3 10.5 12.4 9.7 10.9 13.9
Ear height (cm) Max 115.1 103 65.3 97.6 93.6 113.5 94.9 123.2 117.6
Mean 82.6 85.7 46.3 87.3 74.5 96.2 75.5 86.1 4.2
Min 57.3 71.6 31.8 74.9 61.8 78.6 63.3 68.4 59.7
SD 10.1 6.8 8.6 5.9 8.1 9.5 8.6 12.4 14.1

Note. Underlined numbers indicate the mean, minimum, and SD of the associated row.
a
A negative BLUP value was an artifact of the model fitting all years’ and locations’ combined analysis.

3.1 Variance components and repeatability:
Aflatoxin
Variance components are useful for evaluating sources of vari-
ation in an experiment and allow repeatability to be calcu-
lated. Location × year and the Loc × year × Irr interac-
tion variables for Afppb explained 6.6 and 0%, respectively.
However, when aflatoxin data were transformed the major
sources of variation fell into Loc × year and Loc × year × Irr,
38.3 and 30.1%, respectively. These major sources of varia-
tion for aflatoxin data that, while interesting, are not sources
that a farmer would have much control over. Total genotypic
variation was 3.8% for AfTrans and 3.7% for Afppb in the
full model (all collected data), suggesting genotype was a
872 Crop Science
minor to moderate driver of aflatoxin variation in this exper-
iment; however, it is one a producer can easily manipulate.
This lower genotypic variation also reflects a generally less-
susceptible set of cultivars tested here. Location × year × Irr ×
Gen for transformed aflatoxin explained 3.3% variation, and
this four-way interaction would be useful in selecting reduced
aflatoxin-susceptible genotypes for different environments
such as dryland in Corpus Christi versus irrigated in Col-
lege Station. While mean comparisons showed that atoxigenic
applications reduced aflatoxin, this main effect explained only
2.7% of variation for AfTrans and 4.7% for Afppb. Variation
explained by inoculation was 2.7% and demonstrated that for
this experiment, selecting resistant genotypes would have a
larger effect than applying atoxigenic strains. On transformed
aflatoxin data, Loc × Inoc × Gen and year × Irr × Inoc ×
Gen both explained .2% of the variation. Demonstrating the
effectiveness of atoxigenic control applications depended on
the different locations and genotypes examined. Due to logis-
tics and knowledge of flowering, it was difficult to ensure that
the atoxigenics were applied at the same developmental stage
in each environment, which could have caused some of this
variation. Total variation explained from all variables con-
taining Inoc was 5.42%. Genotype × Inoc and Loc × Inoc
× Gen explained .14% and .16% variation, respectively. This
could suggest that in this experiment, atoxigenics had dif-
ferent effectiveness for genotypes. Comparing the atoxigenic
applications against the check hybrids versus the experimen-
tal hybrids it was observed that the treatment of an atoxigenic
application when applied to the experimental hybrids did sig-
nificantly lower the amount of aflatoxin accrued by up to 38%
(p < .001).
A substantial interaction of 30.1% was seen between Loc
× year × Irr probably brought forth by irrigation or rainfall
timing events during inoculations which may have increased
or decreased the amount of aflatoxin accumulation across dif-
ferent environments. The full model for transformed aflatoxin
data explained 84% variation with only 16% variation remain-
ing as unexplained residual error which is relatively good for
an aflatoxin experiment, perhaps because of the number of
replications tested and the large variations in environments
and genotypes.
Repeatability was used over heritability due to the lack of
family structure in the genotypes. Repeatability is typically
high for flowering and height and lower for yield and aflatox-
ins (Farfan, et al., 2015; Wahl, et al., 2017). Across all envi-
ronments, repeatability for aflatoxin accumulation data was
88% using transformed data down to 72% on nontransformed
raw data (Table 2). Repeatability for aflatoxin accumulation
within individual environments was lower than for all other
important agronomic traits, as expected, but still moderate to
high.
PEKAR ET AL.
3.2 Variance components: grain yield and
secondary agronomic traits
Within the full model, the variation explained for grain yield
was from Loc (55.4%) with only 3.3% explained by the Year.
Only 2.3% of variation was explained by the irrigation regime.
The reduced variation from irrigation could be explained by
rainfall amounts during most seasons that limited moisture
stress from occurring. It was observed that Loc × Year × Irr
had a large effect on AfTrans (30.12%) while Year and Irr
did not have a large effect on Yield. This could be explained
by timing of rainfall or irrigations during the season which
may have had little beneficial effect on plant growth, but a
large increase in Aspergillus flavus growth. Genotype, what
farmers have most control over explained 27.4%, but this was
likely inflated due to some experimental checks with low
yield. Location explained 66.4% variation for PHT, similar to
grain yield, however only 12% of the variation for EHT was
explained by Loc. In EHT the variation was more attributable
to hybrid genotype (17.7%), whereas genotype only explained
7.7% of the variation for PHT.
Flowering variation was mostly explained by Year
(DTS = 77.2%, DTA = 75.7%) indicating that days to flow-
ering was mostly dependent on seasonal variations such as
heat accumulations shared across College Station and Corpus
Christi. Residual error for both DTS and DTA was 1.3 and
1%, respectively. Comparing across all agronomic traits, the
year had the greatest effect on creating experimental variabil-
ity than nearly any other source. This was surprising given
these two environments and the four years were relatively
similar to each other. This is important because environment
(year × location) is what producers have the least ability
to control, suggesting that for yield and agronomic traits,
genetics and management can have limitations in meeting
producer’s needs.
3.3 Correlations
Transformed aflatoxin was significant (p < .01) and nega-
tively correlated with grain yield at r = –.17 while Afppb
correlated negatively with grain yield at r = –.12 (Figure 1).
These negative correlations showed that better-yielding mate-
rial was less aflatoxin-accumulating overall, supporting the
hypothesis that adaptation may be more important than resis-
tance, per se, across diverse elite hybrids. Plant height and
EHT were significant (p < .01) and positively correlated with
yield at r = .73 for PHT and r = .36 for EHT. This positive cor-
relation between plant height and grain yield is very similar to
that found by Farfan et al. (2013) (r = .61) across over 10,000
plots of elite commercial hybrids. Negative correlations
PEKAR ET AL. Crop Science 873

T A B L E 2 Variance components and percent total for traits measured in all years and locations combined for transformed aflatoxin value
(Aftrans), actual measured aflatoxin accumulation in parts per billion (Afppb), yield (T ha-1 ), plant height (PHT), ear height (EHT). Terms defined as
Location (Loc), Year (Year), Irrigation method (Irr), Inoculation method (Inoc), Genotype (Gen), Replication (Rep), Range (Range), and Row (Row)

Model AfTrans Variation Afppb Variation T ha-1 Variation

% % %
Loc 2.82E-11 – – – 1.09E+01 55.42
Loc:Gen 2.04E-02 1.94 1.38E+03 6.13 1.53E-01 .77
Loc:Inoc 2.17E-02 2.06 6.03E-04 – – –
Loc:Inoc:Gen 1.65E-03 .16 4.30E+02 1.91 1.79E-09 –
Loc:Irr – – – – 1.53E-08 –
Loc:Irr:Gen – – 1.94E+01 .09 5.65E-03 .03
Loc:Irr:Inoc – – – – 3.16E-09 –
Loc:Irr:Inoc:Gen 3.67E-11 – – – 6.85E-09 –
Loc:Year 4.01E-01 38.25 1.48E+03 6.56 7.75E-04 –
Loc:Year:Gen 1.58E-08 – 1.72E+03 7.63 1.05E-01 .53
Loc:Year:Inoc 4.02E-03 .38 1.16E+03 5.16 3.83E-10 –
Loc:Year:Inoc:Gen 3.52E-10 – 3.94E+02 1.75 5.29E-09 –
Loc:Year:Irr 3.16E-01 3.12 – – – –
Loc:Year:Irr:Gen 3.43E-02 3.27 1.61E+02 .71 4.12E-02 .21
Loc:Year:Irr:Inoc 1.95E-10 – – – 6.64E-09 –
Loc:Year:Irr:Inoc:Gen 1.09E-09 – 1.74E+02 .77 3.24E-09 –
Gen 3.97E-02 3.78 8.39E+02 3.72 5.40E+00 27.38
Inoc 2.81E-02 2.67 1.07E+03 4.74 – –
Inoc:Gen 1.48E-03 .14 3.72E-05 – 5.72E-10 –
Irr 1.10E-07 – 6.21E+01 .28 4.64E-01 2.35
Irr:Gen – – 5.72E+01 .25 3.22E-09 –
Irr:Inoc 2.36E-10 – – – – –
Irr:Inoc:Gen 1.80E-10 – – – 2.94E-09 –
Range:Loc:Year 1.69E-03 .16 2.20E+02 .97 1.85E-01 .94
Rep:Loc:Year 8.69E-03 .83 4.49E+02 1.99 3.70E-02 .19
Row:Loc:Year 2.70E-04 .03 1.93E-05 – 1.87E-01 .95
Year 2.57E-04 .02 5.37E+03 23.83 6.59E-01 3.34
Year:Inoc – – 6.38E-03 – – –
Year:Inoc:Gen 1.84E-11 – 3.63E-05 – – –
Year:Irr 1.80E-09 – 1.50E+02 .67 7.59E-01 3.85
Year:Irr:Gen 6.73E-10 – 6.41E-05 – 4.52E-06 –
Year:Irr:Inoc 3.74E-11 – – – – –
Year:Irr:Inoc:Gen 1.62E-03 .15 1.44E+02 .64 – –
Residual 1.68E-01 16.02 7.26E+03 32.19 7.97E-01 4.04
Total 1.05E+00 100.00 2.25E+04 100.00 1.97E+01 100.00
R (%) 88.31 71.50 99.42
Model PHT Variation EHT Variation
% %
Loc 9.26E+02 66.44 4.16E+01 11.97
Loc:Gen 1.43E+01 1.03 2.26E+01 6.51
Loc:Inoc 1.05E-08 – – –
Loc:Inoc:Gen 1.18E-07 – – –
Loc:Irr 4.13E+01 2.96 1.08E+01 3.10
(Continues)
874 Crop Science PEKAR ET AL.

TA B L E 2 (Continued)

Model AfTrans Variation Afppb Variation T ha-1 Variation

Loc:Irr:Gen – – – –
Loc:Irr:Inoc – – 1.08E-01 .03
Loc:Irr:Inoc:Gen – – – –
Loc:Year 1.66E+02 11.91 1.94E-06 –
Loc:Year:Gen 6.44E+00 .46 8.99E+00 2.59
Loc:Year:Inoc 1.23E-06 – – –
Loc:Year:Inoc:Gen 4.91E-07 – 5.91E-08 –
Loc:Year:Irr 2.51E-06 – - –
Loc:Year:Irr:Gen 8.23E-08 – 3.09E+00 .89
Loc:Year:Irr:Inoc – – 4.25E-08 –
Loc:Year:Irr:Inoc:Gen 7.92E-08 – 5.66E-07 –
Gen 1.08E+02 7.73 6.13E+01 17.66
Inoc – – – –
Inoc:Gen 3.70E-06 – 1.07E+00 .31
Irr 1.72E-06 – 6.92E-03 –
Irr:Gen 1.16E+00 .08 5.48E-01 .16
Irr:Inoc – – – –
Irr:Inoc:Gen 3.81E-01 .03 – –
Range:Loc:Year 2.13E+01 1.53 6.81E+00 1.96
Rep:Loc:Year 4.46E+00 .32 1.66E+00 .48
Row:Loc:Year 6.99E+00 .50 2.08E+00 .60
Year 1.88E-03 – 7.66E+01 22.07
Year:Inoc 9.33E-02 .01 4.44E-08 –
Year:Inoc:Gen – – – –
Year:Irr 1.31E+01 .94 8.98E+00 2.59
Year:Irr:Gen – – 2.11E-01 .06
Year:Irr:Inoc 1.20E-07 – – –
Year:Irr:Inoc:Gen 1.70E-07 – 6.48E-08 –
Residual 8.45E+01 6.06 1.01E+02 29.01
Total 1.39E+03 100.00 3.47E+02 100.00
R (%) 97.26 94.29
Model DTS Variation DTA Variation
% %
Loc 1.65E+01 15.50 2.10E-05 –
Loc:Gen 1.97E+00 1.85 1.38E+00 1.38
Loc:Inoc 1.10E-08 – 2.98E-03 –
Loc:Inoc:Gen – – – –
Loc:Irr 1.35E-09 – 4.69E-02 .05
Loc:Irr:Gen – – 3.21E-02 .03
Loc:Irr:Inoc – – 1.60E-10 –
Loc:Irr:Inoc:Gen 1.01E-09 – – –
Loc:Year 1.05E-05 – 1.88E+01 18.84
Loc:Year:Gen 2.90E-01 .27 2.63E-01 .26
Loc:Year:Inoc – – 1.64E-10 –
Loc:Year:Inoc:Gen – – 1.31E-02 .01
Loc:Year:Irr 1.39E-01 .13 6.17E-02 .06
(Continues)
PEKAR ET AL. Crop Science 875

TA B L E 2 (Continued)

Model AfTrans Variation Afppb Variation T ha-1 Variation

Loc:Year:Irr:Gen 1.24E-09 – – –
Loc:Year:Irr:Inoc 2.38E-10 – – –
Loc:Year:Irr:Inoc:Gen – – 2.27E-09 –
Gen 3.13E+00 2.94 2.25E+00 2.26
Inoc – – – –
Inoc:Gen 1.87E-02 .02 1.09E-10 –
Irr – – 6.17E-09 –
Irr:Gen 7.32E-02 .07 8.06E-10 –
Irr:Inoc 2.29E-09 – – –
Irr:Inoc:Gen 3.04E-09 – 5.10E-10 –
Range:Loc:Year 3.94E-01 .37 1.72E-01 .17
Rep:Loc:Year 1.01E-01 .09 1.53E-01 .15
Row:Loc:Year 2.19E-01 .21 9.41E-02 .09
Year 8.21E+01 77.19 7.54E+01 75.72
Year:Inoc – – – –
Year:Inoc:Gen – – 6.33E-09 –
Year:Irr – – – –
Year:Irr:Gen 5.25E-02 .05 7.09E-10 –
Year:Irr:Inoc 9.27E-10 – 2.83E-10 –
Year:Irr:Inoc:Gen – – 9.06E-10 –
Residual 1.38E+00 1.30 9.48E-01 .95
Total 1.06E+02 100.00 9.96E+01 100.00
R (%) 98.08 98.00

Note. DTA, days to anthesis; DTS, days to silking.

between PHT and AfTrans also supported the hypothesis that

F I G U R E 1 Phenotypic correlations on all raw data and all years,
across all locations for days to anthesis (DTA), days to silking (DTS),
plant height (PHT), ear height (EHT), yield (T ha-1 ), aflatoxins in ppb
(AFppb), and transformed aflatoxin data (Aftrans)
more vigorous material is less aflatoxin susceptible. Days to
anthesis and DTS were positively correlated and significant
with grain yield (p < .01) at r = .60 and r = .56, respectively.
Later-flowering material accumulating higher-grain yield
is somewhat surprising, because these plants would be
subjected to hotter and more stressful conditions during grain
fill; however, later material is also more likely to be adapted
to sub-tropical environments such as Texas. In contrast, the
commercial checks used in this study were high-yielding
but typically flowered earlier than the experimental hybrids;
therefore, the correlation between later flowering and higher
yield were likely due exclusively to experimental checks.
Genotypic correlations (Figure 2) for transformed afla-
toxin accumulation and flowering time were highly significant
(p < .01) and moderately correlated at r = –.49. This decrease
in aflatoxin accumulation may have resulted from favorable
environmental conditions increasing later in the season but is
more likely due to the adaptation of the later material from
sub-tropical environments (Betran & Isakeit, 2004; Betran,
et al., 2006; Wahl, et al., 2017).
876 Crop Science PEKAR ET AL.

significant for PHT (r = .23*, P < .10). Correlations for PHT

and yield have been observed in several studies and help to
prove that yield is closely correlated with PHT in some grow-
ing regions and can be used as an important predictor in plant
yield (Farfan, et al., 2015; Katsvairo, et al., 2003; Yin, et al.,
2011).

3.4 Recommendations for specific hybrid

lines

The best linear unbiased estimators for each hybrid were

F I G U R E 2 Genotypic correlations on all data and all years,
across all locations as best linear unbiased predictors (BLUPs) for days
to anthesis (DTA), days to silking (DTS), plant height (PHT), ear height
(EHT), yield (T ha−1 ), aflatoxins in ppb (AFppb), and transformed
aflatoxin data (Aftrans). Larger font indicates a larger correlation
In contrast to the phenotypic correlations, genotypic corre-
lations for yield were only somewhat positively correlated and
obtained across all environments, within specific environ-
ments, and under specific management conditions. Over the
entire experiment, hybrids LH195/Tx777, LH195/Tx779,
LH195/{[(CML 326/B104)x(CML285/B104)]−2-2-B-B-B-
B/(CML288/NC300)-B-9-B1-B-B-B-B-B}-B-2-B-B-2-1},
TR8145RR2/{[(CML 326/B104)x(CML285/B104)]−2-2-B-
B-B-B/(CML288/NC300)-B-9-B1-B-B-B-B-B}-B-2-B-B-2-
1}, TR8145RR2/Tx779, and Tx781/Tx777 were estimated to
be the best for both lowest aflatoxin and highest grain yield.
The Texas-by-Texas cross Tx781/Tx777 is especially interest-
ing as it is a completely public hybrid and contains tropical-
derived germplasm in both parents. Increasing the interest in
this hybrid is that it out-yielded the check average for yield in

TA B L E 3 Notable hybrids for yield and low aflatoxin accrual

Log10 (aflatoxin) %Check

Year Pedigree BLUPs %Check average BLUPS average
2014 LH195/{[(CML 326/B104)x(CML285/B104)]−2-2-B-B- 126.93 ± 46.09 93.65 3.1 ± 1.29 –19.06
B-B/(CML288/NC300)-B-9-B1-B-B-B-B-B}-B-2-B-B-
2-1
TR8145RR2/{[(CML 326/B104)x(CML285/B104)]−2-2- 126.04 ± 46.09 92.99 3.27 ± 1.29 –14.62
B-B-B-B/(CML288/NC300)-B-9-B1-B-B-B-B-B}-B-2-
B-B-2-1
2015 NP2643GT/Tx777 135.36 ± 15.87 112 1.22 ± 0.45 –38
Tx779/LH195 134.67 ± 15.9 112 1.12 ± 0.45 –43
Tx781/Tx777 130.48 ± 15.87 108 1.3 ± 0.45 –34
TR8145/Tx777 129.39 ± 15.87 107 1.1 ± 0.45 –44
LH195/Tx777 129.11 ± 15.87 107 1.14 ± 0.45 –42
2016 LH195/Tx777 122.5 ± 27.81 103 3.82 ± 0.95 –10
Tx781/Tx777 122.43 ± 27.81 103 3.81 ± 0.95 –10
LH195/Tx779 121.23 ± 27.98 102 3.71 ± 0.96 –12
NP2643GT/Tx777 121.17 ± 27.81 102 3.87 ± 0.95 –8
GP474GT/Tx780 119.84 ± 27.81 101 3.86 ± 0.95 –9
2017 TR8145RR2/Tx779 102.97 ± 46.91 103 2.48 ± 0.45 –39
GP474?BGTCBLL (glyphosate+liberty link+corn 102.06 ± 46.91 102 3.5 ± 0.46 –13
borer)/Tx779
Tx794/Tx779 97 ± 46.91 97 3.34 ± 0.45 –17
BMP-2/Tx777 96.9 ± 46.91 97 2.74 ± 0.45 –32

Note. BLUP, best linear unbiased predictor.

PEKAR ET AL.
2015 by 107% and in 2016 by 105%. In addition to the high
yield from this hybrid, it accumulated 62% less aflatoxin than
the average checks in 2015, while in 2016 it accumulated
15% less aflatoxin than the average checks. This supports the
overall hypothesis that inbreds and hybrids bred in the central-
South Texas growing region can be better adapted, meaning
they may have both higher grain yield and lower aflatoxin
accumulations than commercial hybrids bred elsewhere that
have better overall adaptation (Ewing, et al., 2019).
Several hybrids did well in both dryland and irrigated
regimes (Table 3). Most notable were the hybrids consisting of
Bolivian material (Tx777 and Tx779); several hybrids tested
shared one of these parents, and again, the Tx781/Tx777
hybrid performed well in both dryland and irrigated regimes.
In previous studies, LH195 consistently produced hybrids
with comparable yield to local commercial checks in the
stressful Texas environments (Murray, et al., 2019). This tester
was used to make several of our hybrids for this study both as
an expired plant cultivar protection line (Foley, 1991) and as
modern-traited inbred. The expired plant cultivar protection
line has shown some differences, as both a line and in hybrid
combination, from the commercially available inbred, though
both look nearly identical.
4 CONCLUSIONS
This study demonstrated the complex interactions between
different technologies that have been proven to reduce afla-
toxin when examined in isolation. Genetics, agronomic man-
agement, and atoxigenic strains all substantially influenced
aflatoxin. While historically studied in isolation, this study
showed that some variation was explained by interactions
between genotype by atoxigenic inoculation in differing envi-
ronments.
Breeding is an important component of integrated pest
management for aflatoxin tolerance. However, complete resis-
tance to aflatoxin accumulation has not been found. It has
been determined that genotypic variation and host resistance
can be exploited by the producer. However small, this can
help producers in choosing lower aflatoxin-accumulating cul-
tivars. Crop management practices can also contribute to an
overall decrease in aflatoxin accumulation in maize. This
study shows that decreases in aflatoxin accumulation can be
achieved through selecting genotypes that are better adapted
in dryland or irrigated regimes, by inoculating with atoxigenic
strains, and possibly through selecting genotypes that are syn-
ergistic with atoxigenic strain applications in reducing toxins.
There were no significant interactions between genotype and
atoxigenic inoculation; however, trends were seen that showed
lowered aflatoxin accrual in certain genotypes.
Twelve hybrids did considerably well compared with
checks for both aflatoxin accumulation and yield, and the
Crop Science 877
majority of entries were of TAMU × TAMU hybrids. This
bias will inherently skew the trend to favor the TAMU ×
TAMU hybrids, which have been selected for adaptation to
the subtropical environment of central Texas that this testing
was conducted in, for resistance to aflatoxin accumulation.
AC K N OW L E D G M E N T S
The authors would like to thank Steven Anderson, David
Rooney, and staff of the Texas A&M AgriLife Research and
Extension Center at Weslaco for their help with this research.
This research was funded by USDA National Institute of Food
and Agriculture Competitive Grant 2014-6804-21836, USDA
Hatch funds, USDA-ARS, Texas A&M AgriLife Research,
the Aflatoxin Mitigation Center of Excellence, and Texas
Corn Producers Board.
AU T H O R C O N T R I B U T I O N S
Jacob J. Pekar: Data curation; Formal analysis; Writing – orig-
inal draft; Writing – review & editing. Seth C. Calder Mur-
ray: Conceptualization; Funding acquisition; Methodology;
Project administration; Supervision; Writing – original draft;
Writing – review & editing. Thomas Isakeit: Methodology;
Supervision; Writing – review & editing. Luke S. Pruter: Data
curation; Writing – review & editing. Nancy J. Wahl: Writ-
ing – review & editing. Michael J. Brewer: Conceptualization;
Methodology; Writing – review & editing.
CONFLICT OF INTEREST
The authors declare there to be no conflict of interest.
ORCID
Jacob J. Pekar https://orcid.org/0000-0002-4559-248X
REFERENCES
Abbas, H., Zablotowicz, R., Horn, B., Phillips, N., Johnson, B., Jin, X.,
& Abel, C. A. (2011). Comparison of major biocontrol strains of non-
aflatoxigenic Aspergillus flavus for the reduction of aflatoxins and
cyclopiazonic acid in maize. Food Additives & Contaminants, 28,
198–208. https://doi.org/10.1080/19440049.2010.544680
Abbas, H. K., Williams, W. P., Windham, G. L., Pringle, H. C., Xie,
W., & Shier, W. T. (2002). Aflatoxin and fumonisin contamination
of commercial corn (Zea mays) hybrids in Mississippi. Journal of
Agricultural and Food Chemistry, 50, 5246–5254. https://doi.org/10.
1021/jf020266k
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting linear
mixed-effects models using {lme4}. Journal of Statistical Software,
67, 1–48. https://doi.org/10.18637/jss.v067.i01
Barrero Farfan, I. D., Murray, S. C., Labar, S., & Pietsch, D. (2013).
A multi-environment trial analysis shows slight grain yield improve-
ment in Texas commercial maize. Field Crop Research, 149, 167–176.
https://doi.org/10.1016/j.fcr.2013.04.017
Betrán, F. J., & Isakeit, T. (2004). Aflatoxin accumulation in maize
hybrids of different maturities. Agronomy Journal, 96, 565–570.
https://doi.org/10.2134/agronj2004.5650
878 Crop Science
Betrán, F. J., Bhatnagar, S., Isakeit, T., Odvody, G., & Mayfield, K.
(2006). Aflatoxin accumulation and associated traits in QPM maize
inbreds and their testcrosses. Euphytica, 152, 247–257. https://doi.
org/10.1007/s10681-006-9207-3
Betrán, F. J., Isakeit, T., & Odvody, G. (2002). Aflatoxin accumulation
of white and yellow maize inbreds in diallel crosses. Crop Science,
42, 1894–1901. https://doi.org/10.2135/cropsci2002.1894
Brown, R., Chen, Z., Warburton, M., Luo, M., Menkir, A., Fakhoury, A.,
& Bhatnagar, D. (2011). Proteins associated with aflatoxin-resistance
are identified and characterized towards candidacy for breeding mark-
ers. Phytopathology, 101, S21–S21.
Brown, R. L., Cotty, P. J., & Cleveland, T. E. (1991). Reduction in afla-
toxin content of maize by atoxigenic strains of Aspergillus flavus.
Journal of Food Protection, 54, 623–626. https://doi.org/10.4315/
0362-028X-54.8.623
Cole, C. B., Bowman, D. T., Bourland, F. M., Caldwell, W. D., Campbell,
B. T., Fraser, D. E., & Weaver, D. B. (2009). Impact of heterozygos-
ity and heterogeneity on cotton lint yield stability. Crop Science, 49,
1577–1585. https://doi.org/10.2135/cropsci2008.08.0450
Drott, M. T., Lazzaro, B. P., Brown, D. L., Carbone, I., & Milgroom,
M. G. (2017). Balancing selection for aflatoxin in Aspergillus flavus
is maintained through interference competition with, and fungivory
by insects. Proceedings of the Royal Society B, 284, 20172408. https:
//doi.org/10.1098/rspb.2017.2408
Ewing, P. M., Runck, B. C., Kono, T. Y. J., & Kantar, M. B. (2019).
The home field advantage of modern plant breeding. PLOS ONE, 14,
e0227079. https://doi.org/10.1371/journal.pone.0227079
Farfan, I. D. B., De La Fuente, G. N., Murray, S. C., Isakeit, T., Huang,
P. C., Warburton, M., Williams, P., Windham, G. L., & Kolomiets, M.
(2015). Genome wide association study for drought, aflatoxin resis-
tance, and important agronomic traits of maize hybrids in the sub-
tropics. Plos One, 10. https://doi.org/10.1371/journal.pone.0117737
Foley, T. J. (1991). Inbred corn line LH195. Google Patents.
Gqaleni, N., Smith, J. E., Lacey, J., & Gettinby, G. (1997). Effects of
temperature, water activity, and incubation time on production of afla-
toxins and cyclopiazonic acid by an isolate of Aspergillus flavus in
surface agar culture. Applied and Environmental Microbiology, 63,
1048–1053. https://doi.org/10.1128/aem.63.3.1048-1053.1997
Henry, W. B., Williams, W. P., Windham, G. L., & Hawkins, L. K.
(2009). Evaluation of maize inbred lines for resistance to Aspergillus
and Fusarium ear rot and mycotoxin accumulation. Agronomy Jour-
nal, 101, 1219–1226. https://doi.org/10.2134/agronj2009.0004
Isakeit, T., Murray, S., & Wilborn, J. (2011). Efficacy of Afla-Guard
(Aspergillus flavus NRRL 21882) to control mycotoxins on corn in
Burleson County, Texas, 2010. Plant Disease Management Reports
5.
Jones, R., Duncan, H., & Hamilton, P. (1981). Planting date, harvest
date, and irrigation effects on infection and aflatoxin production by
Aspergillus flavus in field corn. Phytopathology, 71, 810–816.
Juniper Systems. (1993–2017). Classic GrainGage.
Katsvairo, T. W., Cox, W. J., & Van Es, H. M. (2003). Spatial growth and
nitrogen uptake variability of corn at two nitrogen levels. Agronomy
Journal, 95, 1000–1011. https://doi.org/10.2134/agronj2003.1000
Klich, M. A. (2007). Aspergillus flavus: The major producer of aflatoxin.
Molecular plant pathology, 8, 713–722. https://doi.org/10.1111/j.
1364-3703.2007.00436.x
Kloepper, J. W. (1996). Host specificity in microbe-microbe interactions.
Bioscience, 46, 406–409. https://doi.org/10.2307/1312874
PEKAR ET AL.
Lewis, L., Onsongo, M., Njapau, H., Schurz-Rogers, H., Luber, G.,
Kieszak, S., Nyamongo, J., Backer, L., Dahiye, A. M., Misore, A.,
Decock, K., & Rubin, C. (2005). Aflatoxin contamination of com-
mercial maize products during an outbreak of acute aflatoxicosis in
eastern and central Kenya. Environmental Health Perspectives, 113,
1763–1767. https://doi.org/10.1289/ehp.7998
Li, Z., Coffey, L., Garfin, J., Miller, N. D., White, M. R., Spalding, E.
P., De Leon, N., Kaeppler, S. M., Schnable, P. S., Springer, N. M., &
Hirsch, C. N. (2018). Genotype-by-environment interactions affecting
heterosis in maize. Plos One, 13, e0191321. https://doi.org/10.1371/
journal.pone.0191321
Liu, K., & Wiatrak, P. (2011). Corn (Zea mays L.) plant characteristics
and grain yield response to N fertilization programs in no-tillage sys-
tem. American Journal of Agricultural and Biological Sciences, 6,
279–286. https://doi.org/10.3844/ajabssp.2011.279.286
Mideros, S. X., Windham, G. L., Williams, W. P., & Nelson, R. J.
(2009). Aspergillus flavus biomass in maize estimated by quantitative
real-time polymerase chain reaction is strongly correlated with afla-
toxin concentration. Plant Disease, 93, 1163–1170. https://doi.org/10.
1094/PDIS-93-11-1163
Murray, S. C., Mayfield, K., Pekar, J., Brown, P., Lorenz, A., Isakeit,
T., Odvody, G., Xu, W., & Betran, J. (2019). Tx741, Tx777, Tx779,
Tx780, and Tx782 inbred maize lines for yield and southern United
States stress adaptation. Journal of Plant Registration. https://doi.org/
10.3198/jpr2017.07.0044crp
Payne, G. A., & Widstrom, N. W. (1992). Aflatoxin in maize. Criti-
cal Reviews in Plant Sciences, 10, 423–440. https://doi.org/10.1080/
07352689209382320
Pekar, J. J., Murray, S. C., Isakeit, T. S., Scully, B. T., Guo, B., Knoll, J.
E., Ni, X., Abbas, H. K., Williams, P., & Xu, W. (2019). Evaluation
of elite maize inbred lines for reduced Aspergillus flavus infection,
aflatoxin accumulation, and agronomic traits. Crop Science. https://
doi.org/10.2135/cropsci2019.04.0206
R Development Core Team. (2010). R: A language and environment for
statistical computing. R Foundation for Statistical Computing.
Robens, J., & Cardwell, K. (2003). The costs of mycotoxin management
to the USA: Management of aflatoxins in the United States. Journal
of Toxicology: Toxin Reviews, 22, 139–152.
SAS Institute Inc. (1989–2019). JMP 14. SAS Institute Inc.
Schnell, F. W., & Becker, H. C. (1986). Yield and yield stability in
a balanced system of widely differing population structures in Zea
mays L. Plant Breeding, 97, 30–38. https://doi.org/10.1111/j.1439-
0523.1986.tb01298.x
Smith, K. P., & Goodman, R. M. (1999). Host variation for interactions
with beneficial plant-associated microbes. Annual Review of Phy-
topathology, 37, 473–491. https://doi.org/10.1146/annurev.phyto.37.
1.473
Stoloff, L., Van Egmond, H., & Park, D. (1991). Rationales for the estab-
lishment of limits and regulations for mycotoxins. Food Additives &
Contaminants, 8, 213–221.
Syngenta Crop Protection. (2017). Afla-Guard GR biocontrol agent.
Tucker, D., Jr., Trevathan, L., King, S., & Scott, G. (1986). Effects of
four inoculation techniques on infection and aflatoxin concentration
of resistant and susceptible corn hybrids inoculated with Aspergillus
flavus. Phytopathology, 76, 290–293.
Wahl, N., Murray, S. C., Isakeit, T., Krakowsky, M., Windham, G.
L., Williams, W. P., Guo, B., Ni, X., Knoll, J., Xu, W., Scully, B.,
Mayfield, K., & Betran, J. (2017). Identification of resistance to
PEKAR ET AL.
aflatoxin accumulation and yield potential in maize hybrids in the
Southeast Regional Aflatoxin Trials (SERAT). Crop Science, 57(1),
202–215. https://doi.org/10.2135/cropsci2016.06.0519
Wahl, N. J., Murray, S. C., Zhang, H.-B., Zhang, M., Dickens, C. M.,
& Isakeit, T. S. (2019). Maize kernel development stage the primary
factor in differential gene expression in response to two methods of
field inoculation with Aspergillus flavus. bioRxiv. 617241. https://doi.
org/10.1101/617241
Widstrom, N. W., Guo, B. Z., & Wilson, D. M. (2003). Integration of crop
management and genetics for control of preharvest aflatoxin contam-
ination of corn. Journal of Toxicology: Toxin Reviews, 22, 195–223.
https://doi.org/10.1081/TXR-120024092
Williams, W. P. (2006). Breeding for resistance to aflatoxin accumu-
lation in maize. Mycotoxin Research, 22, 27–32. https://doi.org/10.
1007/BF02954554
Williams, W. P., Alpe, M. N., Windham, G. L., Ozkan, S., & Mylroie,
J. E. (2013). Comparison of two inoculation methods for evaluat-
ing maize for resistance to Aspergillus flavus infection and aflatoxin
accumulation. International Journal of Agronomy, 972316. https://
doi.org/10.1155/2013/972316
Williams, W. P., Buckley, P. M., & Windham, G. L. (2002). Southwestern
corn borer (Lepidoptera: Crambidae) damage and aflatoxin accumu-
lation in maize. Journal of Economic Entomology, 95(5), 1049–1053.
https://doi.org/10.1603/0022-0493-95.5.1049
Williams, W., & Windham, G. (2015). Aflatoxin accumulation in a
maize diallel cross. Agriculture, 5(2), 344–352. https://doi.org/10.
3390/agriculture5020344
Crop Science 879
Windham, G. L., Williams, W. P., Hawkins, L. K., & Brooks,
T. D. (2009). Effect of Aspergillus flavus inoculation methods
and environmental conditions on aflatoxin accumulation in corn
hybrids. Toxin Reviews, 28(2–3), 70–78. https://doi.org/10.1080/
15569540802450037
Wu, F. (2015). Global impacts of aflatoxin in maize: Trade and human
health. World Mycotoxin Journal, 8(2), 137–142. https://doi.org/10.
3920/WMJ2014.1737
Wu, F., & Guclu, H. (2012). Aflatoxin regulations in a network of global
maize trade. Plos One, 7, e45151. https://doi.org/10.1371/journal.
pone.0045151
Yin, X., McClure, M. A., Jaja, N., Tyler, D. D., & Hayes, R. M. (2011).
In-season prediction of corn yield using plant height under major pro-
duction systems. Agronomy Journal, 103, 923–929. https://doi.org/
10.2134/agronj2010.0450
How to cite this article: Pekar, J. J., Murray, S. C.,
Isakeit, T., Pruter, L. S., Wahl, N. J., & Brewer, M. J.
(2022). Control of aflatoxin using atoxigenic strains
and irrigation management is complicated by maize
hybrid diversity. Crop Science, 62, 867–879.
https://doi.org/10.1002/csc2.20710